Nitrogen (N) mineralised from soil organic matter can be an important source of N for crop uptake, particularly following cultivation of pastures. Difficulty in predicting the contribution of mineralisation continues to be a serious obstacle to implementating best management practices for fertiliser N. We evaluated biological tests (i.e. net N mineralised in a 28-day aerobic incubation and anaerobically mineralisable N, AMN) and chemical tests (ammonium-N hydrolysis in hot 2 m KCl) as predictors of N supply to a glasshouse-grown oat (Avena sativa L.) crop. The oat plants were grown to maturity without added N on 30 soils representing a range of management histories, including soils collected from long-term pastures and intensive arable cropping sites. The majority (average 83%) of the N accumulated in grain and straw was mineralised N. Plant N derived from mineralisation (PNDM), estimated by subtracting soil mineral N at sowing from N uptake, was generally higher for long-term pasture soils (mean 82 mg/kg, n = 9) than for long-term arable soils (mean 48 mg/kg, n = 9). The 2 measures of N mineralisation were not closely related [R² = 0.11 (0.37*** when one outlying observation was omitted)], indicating that aerobic and anaerobic assays can give quite different N fertility rankings. Aerobically mineralisable N was the best predictor of PNDM (R² = 0.79***). The ratio of CO₂-C evolved to net N mineralised in the aerobic incubation was highly variable (e.g. mean of 13.6 for pasture soils v. 7.5 for long-term arable soils), likely due to differences in N immobilisation. The correlations of AMN (R² = 0.32**) and hot KCl N (R² = 0.24**) with PNDM were not much better than that between total soil N and PNDM (R² = 0.16*), suggesting that these tests would not provide reliable estimates of N mineralisation potential in soils with diverse management histories.